Spatially-Adaptive Filter Units for Deep Neural Networks
arXiv:1711.11473
Abstract
Classical deep convolutional networks increase receptive field size by either gradual resolution reduction or application of hand-crafted dilated convolutions to prevent increase in the number of parameters. In this paper we propose a novel displaced aggregation unit (DAU) that does not require hand-crafting. In contrast to classical filters with units (pixels) placed on a fixed regular grid, the displacement of the DAUs are learned, which enables filters to spatially-adapt their receptive field to a given problem. We extensively demonstrate the strength of DAUs on a classification and semantic segmentation tasks. Compared to ConvNets with regular filter, ConvNets with DAUs achieve comparable performance at faster convergence and up to 3-times reduction in parameters. Furthermore, DAUs allow us to study deep networks from novel perspectives. We study spatial distributions of DAU filters and analyze the number of parameters allocated for spatial coverage in a filter.
arXiv commentAccepted to Computer Vision and Pattern Recognition 2018
References in corpus (12)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Rethinking Atrous Convolution for Semantic Image Segmentation
- Deep Residual Learning for Image Recognition
- Caffe: Convolutional Architecture for Fast Feature Embedding
- Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
- Multi-Scale Context Aggregation by Dilated Convolutions
- Fully Convolutional Networks for Semantic Segmentation
- DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
- Deformable Convolutional Networks
- Aggregated Residual Transformations for Deep Neural Networks
- Dilated Residual Networks
- Towards Deep Compositional Networks